I am reading the John Purcell's book, and finding it very useful.

I need to set my session at the best way to improve my work. Right now I am editing a dialogue, and I will have to mix, too, a low budget movie. Audio has been recorded directly on dslr, so I have had just one omf track mono of dialogue, and many stereo music tracks to mix.

I don't have any external hardware mixer, all my job is based on the DAW (PT10) (audio interface has 4 channel outputs).

In John Purcell's book I read, that he uses 12-14 tracks for each scene and never uses the same tracks on adjacent scenes, and he does a checkerboard about angle and/or about character.

(I come from music, and there you normally set each track with one and only one instrument).

I get confused in about the how to save tracks, but if I want to use some plug in, am is assuming to have same plug in on all those scenes based on those tracks?


4 Answers 4


I agree the Purcell book is a great start but the typical indie film budget requires something a bit more streamlined. For a quick mix, I try to keep my session limited to 6 or 8 dialog tracks. I checkerboard regions for characters, scenes and any big change in background tone. I don't have a completely new set of tracks for each scene though. Then I route everything to a dialog track with light compression, WNS, C4 and an EQ. I automate parameters on all these plugins by looping a section, suspending automation, dialing in everything the way I want it then either writing automation to selection (HD or CPTK only) or putting the track into automation write mode and playing through that section. Occasionally I'll have dedicated tracks for phones or Naration or monsters or anything that needs an especially unique set of plugins. I hope that all made sense.


And do not forget about the destructive effects in the audiosuite in PT.

The following setup is for tv and Internet ads.

My setup is 4 dialogue tracks routed to one bus, 3 atmospheric/roomtone/music tracks routed to One bus and 5 tracks for misc and fx going to One bus and One Track for lfe. Beside these, i use 4 send bus for convolution reverbs with different predelays, parallel compression and parallel saturation.

I Add subtractive Eq, noise surpression, gain, pan and compression to the samples via the audiosuite. In each buss i use a linMB multiband compressor folliwed by a fairchild compressor. So i can "push" stuff into the compressor by adjusting the gains. This is not the "correct Way" but it works for me.

Via automation i controll the send amounts. Slight amounts of parallel compression and saturation helps glueing the Sounds together. In the master i use a braunworx controll (controls stereowidth) lineq, linMB and a kramer tapesimulation which smooth out the Sounds and magical increases the width of high frequencys. At the end of the Chan i use the izotope Ozons 5 maximizer to controll undetected peaks(i do Not Squash the Sounds by increasing the gain i simply use it as a safety device to Remove peaks and dc)

After the Signal Chain is a visLm meter to meassure the LUFS for TV productions.

I hope this helps!

Edit: to process the seperate dialogue samples, i use two opto compressors (One compression and One limiting). A waves deesser if needed, waves x Noise ans noise surpression and lin phase eq, waves Vocals rider and a large amounts of editing and volume autonation. Its realy important to learn when and how to apply those plugins and when editing is enough.

  • Audiosuite does not perform destructive editing or effects. Audiosuite only changes clips in the timeline, not the original source file, therefore it is not destructive.
    – Mark
    Apr 25, 2019 at 23:04

Good to hear that you are reading and applying Purcells' methods! It's not the only way to do dialogue editing, but it's a great start. Regarding the use of plugins and checkerboarding. The idea of the checkerboard is mainly handy if you have different locations. A scene in a bar requires different eq and reverb than a street scene. That's why you use different tracks and different routings.

In ProTools you can route all dialogue tracks to an auxiliary with a plugin inserted for overall sound enhancements (like a reverb, limiter or compressor).

If you use PT10 is there really a need to save tracks. Or better said do you really need that many tracks for music mixing?


Honestly, I loved John's Purcell's book, but track layout from this book is a little bit different than what I've seen in real life done by prof. dx editors. Well, actually, it depends on mixers to, what track layout they like.

4 Dial tracks is enough for production audio, 2 for a scene, than checkerboard to another pair of tracks, it's not necessary to cut every character for different track. Cut for problem in a speciffic angle, keep it as simple and as logical as possible. Than you may want to have a pfx track, 2 adr track, 2 fill tracks. And make dial tracks work on they're own - you don't need to have a seperate track for room tone only - you can put it on dial tracks, use fill tracks (room tones track) only underneath adr.

Good approach is to go through a scene first and put a marker on every single problem you hear - pops, clicks, wind noise, cloth noise, noisy, low lines, stuff you would like to replace with different takes. The fastest way to look for a different take is to grab it from different angle/take from a regions around it.

And don't put plugins, noise reduction, and stuff, focus on clear and smooth edit first. While mixing you can automate the plugins for any region you want, you will req. a lot of different eq settings from region to region.


  • 3
    I respectfully disagree about the amount of tracks - I've cut quite a lot of dialogue and find that at least 7-8 are necessary for a properly organized layout (for the dub stage to do what they do best without hindrance), especially on scenes which require extensively creative editing to make them work. I know that if I were to deliver a 4-track DX edit I'd get laughed off a stage pretty darn quick (that's best case scenario). May 16, 2013 at 5:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.